Model comparison
Gemma 3 4B vs Inkling-Small
Inkling-Small is the stronger model overall, scoring 46.5 to 28.1 on the Noometry Index. Gemma 3 4B costs 13× less per token, which makes it the better buy when Inkling-Small's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
GoogleSummary
- They share 17 benchmarks with published results for both. Gemma 3 4B scores higher in 0 categories and Inkling-Small in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Inkling-Small leads 48.2 to 11.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 7.5% for Gemma 3 4B and 90% for Inkling-Small.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $0.45 / $1.20 for Inkling-Small.
- Inkling-Small accepts more context: 524K tokens versus 131K.
Side by side
| Gemma 3 4B | Inkling-Small | |
|---|---|---|
| Provider | Thinking Machines Lab | |
| Noometry Index | 28.1 | 46.5 |
| Released | 2025-03-12 | 2026-07-15 |
| Weights | Open | Open |
| Context window | 131K | 524K |
| Max output | 4K | 1.05M |
| Input $ / M tokens | $0.04 | $0.45 |
| Output $ / M tokens | $0.08 | $1.20 |
| Results tracked | 22 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling-Small leads
Gemma 3 4B: 35.9 (#215), Inkling-Small: 43.6 (#85)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| LMArena Coding | 1230 | 1451 |
| LMArena WebDev | — | 1409 |
| SciCode | — | 48.7% |
Agentic & Tool Use Not comparable
Gemma 3 4B: 20.9 (#142), Inkling-Small: —
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | 19.6% | — |
Reasoning Inkling-Small leads
Gemma 3 4B: 13.2 (#335), Inkling-Small: 38.6 (#63)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| Chess Puzzles | 0% | 18% |
| LMArena Hard Prompts | 1253 | 1423 |
| Epoch Capabilities Index | 116.02 | 150.15 |
| ARC-AGI-2 | — | 40.1% |
| Kagi LLM Benchmark | 25.2% | — |
| ARC-AGI-1 | — | 84% |
| CritPt | — | 8.3% |
| Mystery Game Puzzles | — | 6% |
| DTBench | 50.9% | — |
| LMCA | 2.8% | — |
Math Inkling-Small leads
Gemma 3 4B: 16.8 (#292), Inkling-Small: 45.1 (#77)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 7.5% | 90% |
| LMArena Math | 1239 | 1459 |
| FrontierMath (Tiers 1-3) | — | 46.3% |
| FrontierMath Tier 4 | — | 17.1% |
| ProofBench | — | 6% |
Knowledge Inkling-Small leads
Gemma 3 4B: 11.8 (#299), Inkling-Small: 48.2 (#77)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| GPQA Diamond | 23.2% | 88.5% |
| LMArena Expert | 1223 | 1442 |
| SimpleQA Verified | — | 19.1% |
| Vectara Hallucination Rate | 6.4% | — |
Multimodal Not comparable
Gemma 3 4B: —, Inkling-Small: 39.1 (#62)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| LMArena Vision | — | 1235 |
Multilingual Inkling-Small leads
Gemma 3 4B: 42.5 (#194), Inkling-Small: 51.7 (#104)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1273 | 1402 |
| LMArena German | 1281 | 1405 |
| LMArena Russian | 1294 | 1391 |
| LMArena Chinese | — | 1465 |
| LMArena French | — | 1436 |
| LMArena Japanese | — | 1405 |
| LMArena Korean | — | 1363 |
| LMArena Spanish | — | 1428 |
Instruction Following Inkling-Small leads
Gemma 3 4B: 65.2 (#225), Inkling-Small: 73.8 (#114)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1239 | 1399 |
Long Context Inkling-Small leads
Gemma 3 4B: 38.7 (#194), Inkling-Small: 42.7 (#118)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1273 | 1401 |
Writing & Preference Inkling-Small leads
Gemma 3 4B: 42.0 (#239), Inkling-Small: 59.6 (#107)
| Benchmark | Gemma 3 4B | Inkling-Small |
|---|---|---|
| LMArena Text | 1291 | 1414 |
| LMArena Creative Writing | 1271 | 1331 |
| EQ-Bench Creative Writing | 1068 | 1491 |
| LMArena Multi-Turn | 1255 | 1418 |
Frequently asked questions
Is Gemma 3 4B better than Inkling-Small?
Inkling-Small is the stronger model overall, scoring 46.5 to 28.1 on the Noometry Index. Gemma 3 4B costs 13× less per token, which makes it the better buy when Inkling-Small's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 4B or Inkling-Small?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; Inkling-Small lists at $0.45 and $1.20.
Is Gemma 3 4B or Inkling-Small better for coding?
Inkling-Small scores higher on coding benchmarks: 43.6 versus 35.9 in the Noometry coding category.
Which has the bigger context window?
Inkling-Small does, with 524K tokens against 131K.
How many benchmarks do Gemma 3 4B and Inkling-Small share?
17 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Inkling-Small has 33.